Neel Kant

Actually the `Team` abstraction in #206 would be right way to do this. The tasks themselves probably shouldn't have agent-specific instructions. Instead a team config should delegate responsibility and/or set...

Good catch. I think when I got started I casually changed it and didn't bother committing it.

See #207

Full utilization of the A2A spec and sdk suggests having each agent run as an individual server as opposed to being merely clients which interface with a central server. This...

There are already few shot examples of programs in the prompt which come from here: https://github.com/JackHopkins/factorio-learning-environment/blob/4a960edc804142e9de22f7c8a665fb0becccd331/fle/env/tools/agent.md?plain=1#L4 I don't think I've seen examples of helper functions in the prompt, so we...

I tried sorting the agent actions in the codebase and this is what I have: ## core actions (must include) - craft_item - extract_item - harvest_resource - insert_item - launch_rocket...

Referring to @MortenTobiasNielsen 's API proposal: https://github.com/MortenTobiasNielsen/FLE-API-specification/issues/5 The core actions in FLE fully cover the core actions in the API proposal. To leverage human gameplay we will likely want to...

I think the next step is for us to work together on the data collected from @hrshtt 's process to analyze if there are any gaps and how to convert...

I think the state description makes sense. What is the advantage of having actions segmented by domains? If we want to describe a trajectory as a list of action strings...

> if the true distribution has multiple-modes (map ops vs. inventory ops vs. research ops), modelling it as one flat categorical wastes bits. I don't imagine us encoding actions as...